On the evening of May 1, 2026, a cluster of supercell thunderstorms swept across central Oklahoma, spinning up tornadoes that forced tens of thousands of residents into shelters. The warnings went out, but at several National Weather Service forecast offices across the southern Plains, the meteorologists issuing them were doing the jobs of two or three people at once. One year of firings and early retirements has cut the agency’s scientific workforce by roughly 15%, and the consequences are landing squarely on the communities that depend on timely tornado warnings.
A workforce hollowed out before peak season
A New York Times investigation published May 5 found that the NWS scientific staff, which had numbered more than 2,500, shrank by about 15% over the prior year. The losses were not the result of normal turnover. They came from a combination of federal firings and an early-retirement wave that accelerated after a government-wide hiring freeze took hold in 2025.
The damage is concentrated where it matters most: local forecast offices. The NWS operates 122 of them nationwide, and an employee-compiled vacancy tracker, independently verified by the Associated Press in May 2026 through direct calls, staff-list cross-checks, and interviews, shows that 55 offices now carry vacancy rates of 20% or higher. These are the offices responsible for tornado watches, flash-flood warnings, and winter-storm advisories. When a fifth of the desks sit empty during a tornado outbreak, the remaining forecasters absorb every extra radar sector, every additional phone call to emergency managers, and every warning that has to be drafted, reviewed, and issued in minutes.
The squeeze extends into aviation safety. A Government Accountability Office report documented that NWS staffing for FAA center-weather-service-unit meteorologists had fallen to just 69 as of June 2025, a decline the GAO attributed to the hiring freeze and a deferred-resignation program. Those meteorologists coordinate rerouting decisions with air traffic controllers when severe thunderstorms or tornadoes threaten busy airspace. Fewer of them means slower coordination during fast-moving events, and the number may have dropped further in the months since the audit.
NOAA says it is hiring, but the math is daunting
NOAA, the NWS’s parent agency, has acknowledged the problem in stages. After multiple rounds of cuts, officials announced a plan to hire for what they described as mission-critical NWS roles, and the agency received authorization to bring on new staff ahead of the 2026 Atlantic hurricane season. But that green light arrived after the losses had already taken hold.
The timeline for recovery is steep. Training a new forecast meteorologist to issue warnings independently typically takes six months or longer, and the experienced staff who left through early retirement carried institutional knowledge that no onboarding manual can replicate: local climate patterns, radar quirks, relationships with county emergency managers. NOAA has not specified how many positions it intends to restore, which offices will receive new hires first, or when those meteorologists will be fully operational.
“We are stretched to the breaking point,” one NWS forecaster at a southern Plains office told the Associated Press, describing shifts where a single meteorologist monitors severe weather across multiple counties that would normally be split among two or three staff members. Dan Sobien, president of the National Weather Service Employees Organization, told the New York Times that the cuts have left offices “unable to maintain the level of service the public expects and deserves during severe weather events.”
What the data shows and what it does not
No publicly available institutional data directly links the staffing reductions to specific forecasting failures during recent tornado outbreaks. The vacancy figures are well documented, but the NWS has not released internal metrics comparing warning lead times or forecast accuracy before and after the cuts. Without that data, the connection between fewer scientists and slower warnings rests on workload logic and employee accounts rather than measured performance gaps.
That gap matters. A forecast office in Norman, Oklahoma, that lost three of its twelve meteorologists faces a fundamentally different reality than a coastal office in Maine that lost one of fifteen. The 15% topline figure, drawn from the agency’s own roster data and reported by the Times, is reliable at the national level but does not reveal where the pain is worst. The AP-verified vacancy tracker fills part of that hole by identifying the 55 hardest-hit offices, yet it captures a snapshot rather than a trend line showing how conditions have shifted week to week as storm season intensifies.
There is also an open question about whether the 15% figure captures the full scope of the problem. Forecast offices rely on support staff, IT specialists, and administrative personnel whose numbers may have shifted independently of the scientific workforce. If those roles were cut too, the operational burden on remaining meteorologists would be even heavier than the headline number suggests.
What would settle the debate is the kind of before-and-after performance analysis the NWS has so far kept internal: tornado warning lead times, false-alarm rates, and verification scores for severe storms, broken out by office and year. Congressional oversight committees and press organizations have pressed for that data, but as of late May 2026, it has not been made public. Until it is, the relationship between workforce losses and warning quality remains a matter of informed inference, not documented fact.
The May 1 Oklahoma outbreak exposed the cracks
The May 1, 2026, tornado outbreak in central Oklahoma offered a real-time illustration of what staffing shortfalls look like in practice. According to AP reporting, at least two NWS offices covering the affected area were operating with fewer than 75% of their normal forecasting staff that evening. Warnings were issued for every confirmed tornado, and no community went entirely without notice. But employees described the experience as a scramble in which forecasters toggled between multiple warning polygons, delayed routine forecast updates, and deferred coordination calls with county emergency managers until after the immediate threat had passed.
The outbreak did not produce a clear-cut missed warning, and the NWS has not acknowledged any degradation in its performance that night. What it did produce, according to forecasters who spoke to the AP and the Times, was a situation in which the margin between a timely warning and a late one narrowed to seconds rather than minutes. In tornado forecasting, that margin is the difference between a family reaching a shelter and a family caught in the open.
For people living in tornado-prone regions, the verified facts support a serious but measured concern. The NWS still operates a nationwide network of forecast offices staffed by trained meteorologists, and there is no evidence that the warning system has collapsed. But the combination of a 15% reduction in scientific staff, widespread vacancies at local offices, and thinner aviation-weather coverage means the system is running with less cushion precisely when severe weather across the central United States is intensifying.
In practical terms, that shifts more weight onto individual and local preparedness. Residents cannot control how quickly NOAA fills vacancies, but they can make sure wireless emergency alerts are enabled on their phones, maintain a weather radio, and pay close attention to local forecasts on days when the Storm Prediction Center highlights elevated risk. Emergency managers may need to plan for scenarios in which warning lead times are shorter than usual or in which a single forecaster is juggling multiple warning areas simultaneously.
How the NWS rebuilds will shape the next tornado season
The staffing numbers are not abstract. Every vacancy at a forecast office represents a shift that someone else has to cover, a radar screen that gets checked less frequently, or a phone call to a county emergency manager that gets made a few minutes later than it otherwise would. In tornado forecasting, minutes are the unit of survival. The NWS has operated with thin margins before, but the scale of the current losses, hitting nearly half of all forecast offices above the 20% vacancy threshold during peak severe-weather season, is without recent precedent. How the agency rebuilds, and how quickly, will determine whether the spring of 2026 is remembered as a warning or a turning point.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
